近红外光谱技术快速鉴别碱水浸泡鸡肉的研究  

Study on rapid identification of alkali-soaked chicken meat by near-infrared spectroscopy

作　　者：章明[1] 樊艳凤[1] 沈啸 唐修君[1] 陆俊贤[1] 高玉时[1] ZHANG Ming;FAN Yanfeng;SHEN Xiao;TANG Xiujun;LU Junxian;GAO Yushi(Key Laboratory for Poultry Genetics and Breeding of Jiangsu Province/Jiangsu Institute of Poultry Sciences,Yangzhou 225125,China)

机构地区：[1]江苏省家禽科学研究所/江苏省家禽遗传育种重点实验室,江苏扬州225125

出　　处：《扬州大学学报（农业与生命科学版）》2024年第1期101-105,共5页Journal of Yangzhou University：Agricultural and Life Science Edition

基　　金：江苏省农业科技自主创新资金项目[CX(21)2011];扬州市社会发展项目(YZ2022090)。

摘　　要：采用近红外光谱分析技术并结合主成分分析法,建立碱水浸泡鸡肉快速鉴别模型。原始光谱经平滑、多元散射校正和一阶导数等预处理后进行系统主成分分析,结果显示正常鸡肉和碱水浸泡鸡肉能得到清晰的分类。通过不同光谱预处理方法,采用簇类的独立软模式法建立分类模型,结果显示通过对样品光谱采用7点卷积平滑方法预处理,分类模型综合正确率最高,校正回判正确率在77.08%~100%之间,预测正确率在81.25%~100%之间。综上,采用近红外光谱分析技术对碱水浸泡鸡肉进行快速鉴别可行。A rapid identification model of alkaline-soaked chicken meat was established by using near-infrared spectral analysis combined with principal component analysis. The raw spectra were pretreated by smoothing, multiple scattering correction and first-order derivatives, and then subjected to systematic principal component analysis. The results showed that normal chicken meat and alkaline-soaked chicken meat could be clearly classified. The classification models were established by different spectral preprocessing methods using the soft independent modeling method. The results showed that by preprocessing the sample spectra with the 7-point savizkg golag method, the classification models had the highest correct rate. The correct rate of correction back judgment was 77.08%-100% and the correct rate of prediction was 81.25%-100%. The study showed that the NIR spectral analysis technique is feasible for the rapid identification of alkali-soaked chicken meat.

关 键 词：碱水浸泡鸡肉 近红外光谱技术 主成分分析 

分 类 号：S831[农业科学—畜牧学] TS207.3[农业科学—畜牧兽医]